The use of robotic process automation in energy trading

Although not yet widely implemented, it is quickly becoming apparent that significant opportunities exist for the use of robotic process automation (RPA) in energy trading to reduce costs and improve organizational effectiveness. RPA involves the leveraging of digital techniques that can perform simple and repetitive clerical activities, such as data entry, purchase order issuance and the submission of compliance forms. Given the extensive volume of information processed in energy trading operations, RPA thus offers the potential for significantly reducing errors and costs across the entire enterprise, but especially in back-office functions.

The “bots” of robotic process automation

At the heart of RPA are robots — but digital ones instead of physical ones. In RPA, “bots” are introduced into existing software applications in order to capture, interpret and process information that is currently manipulated in well-defined tasks by human workers. These bots interface with a company’s information systems in ways that mimic a clerk: logging in, accumulating specific pieces of data, processing it using structured rules, sending the resulting work to another destination and then logging out.

The benefits of RPA are especially important for companies relying on legacy software applications, many of which cannot be configured to coordinate well with other applications, in large part because trained support is no longer offered for the programs. Oftentimes in these situations, users are required to manually migrate information from one application to another, which is time-consuming, error-prone and hence costly. RPA can automate such transfers.

To illustrate the value-creation potential of RPA, the London School of Economics and Political Science produced a case study profiling RPA adoption at a large European utility. At this utility, bots have automated about 25 back-office processes relating to “meter management, customer billing, account management, consumption management, segmentation and exception processing.” About 300 bots (under the supervision of two employees) have replaced a staff of about 600 in processing about 1 million transactions each month. Against the costs of implementation, the RPA initiative achieved a payback of initial investment within 12 months, and produces an annual ROI of about 200 percent.

As stated by the Institute for Robotic Process Automation and Artificial Intelligence, “any company that uses labor on a large scale for general knowledge process work, where people are performing high-volume, highly transactional process functions, will boost their capabilities and save money and time with robotic process automation software.”

Value in energy trading

Clearly, the above description applies to companies involved in energy trading. Fundamentally, an energy trading operation is an accumulation of a very large number of transactions — both purchases and sales — of energy. Several data elements (product specification, price, volume, delivery time and delivery location) must be stored for each one of these transactions, as they will subsequently be processed in various ways for a number of purposes, including:

• Financial settlement — Buyers must be billed, sellers must be paid and the in-house accounting group must be informed of cash flows in and out of the company.
• Scheduling — After a transaction has been conducted with a counterparty, the physical delivery of the energy must be coordinated with the organizations responsible for nominations and logistics.
• Risk management — After each transaction, the net position of the trading firm changes, and various parties within the company must periodically assess value-at-risk relative to company limits.
• Analytics — As the portfolio changes with each transaction, analytic teams continually reassess the economics and financial implications of available trades that can be made.

Unsurprisingly, energy trading operations employ sizable staffs of clerks to manage the multidirectional and multidimensional flow of data across the organization. Much of this work is rote, low-value and unfulfilling — and it takes time. RPA can significantly reduce these factors. As illustrated in this case study by Greysoft of a bond-trading operation, RPA can reduce processing times by about 50 percent while saving millions of dollars in operating expenditures.

Simultaneous to reducing costs, increasing speed and improving quality, the adoption of RPA can allow employees to focus their intellectual capabilities to more creative value-enhancing purposes, which cannot easily be performed by digital bots.

Implications for companies in energy trading

Although some elements of RPA have been practiced for decades, the breadth and depth of its applicability have been substantially expanded by recent improvements in artificial intelligence (AI). The ecosystem providing RPA solutions is still quite fluid and evolving.

Most companies involved in energy trading — indeed, most companies in all industries — have yet to capture even a significant fraction of the potential benefits that RPA can generate. To date, there’s no proven path for implementing RPA enterprise-wide in an energy trading firm.

Thus, while there is considerable value creation opportunity afforded by RPA in energy trading, the lack of a turnkey solution already dominating the marketplace means that firms seeking to capture the opportunity will need to undertake a dedicated effort involving substantial planning. Any investigation must address the following questions:

• Which processes are so labor-intensive and so error-ridden that the benefits of implementing RPA will quickly outweigh its costs? Ideally, a company seeking to adopt RPA will first utilize it in high-value applications and then consider rolling RPA out incrementally to other applications once the most fruitful sources of value have been captured. Inevitably, not every source of inefficiency caused by manual data intervention will be sufficiently problematic to merit application of RPA.
• How can a foundational capability for RPA be instituted? Given the risks inherent in trying to address all possible RPA use cases in an energy trading operation, it will be important to build an organizational skill for RPA that will allow its use to be tailored and modified as the business evolves and grows. Because unforeseen consequences of any approach based on AI remain a possibility, the internal RPA team needs to be vigilant to ensure that the bots don’t result in processes going awry. Since it’s always better to tackle problem areas one by one rather than attempt to achieve an all-encompassing enterprise solution in one massive undertaking, a company pursuing RPA is well-served to develop in-house abilities to enable continual optimization of RPA use.
• How will the adoption of RPA affect the organization’s workforce? For the company’s employee base, RPA is somewhat of a double-edged sword. On the plus side, RPA promises the elimination of dull tasks. However, RPA might threaten certain employees who are comfortable with their current activities. Employee morale and retention are at stake. Conversely, the implementation of RPA is likely to also change the profile of the most desirable recruits for data processing and management roles.

Though not without costs or the potential for adverse consequences, robotic process automation offers great promise to energy trading operations. Although trading firms can probably survive for quite a while without aggressively adopting RPA, competitive disadvantages may grow over time and become significant if the benefits of RPA aren’t pursued and captured. As with all emerging areas of business innovation, companies involved in energy trading should begin exploring how to implement RPA, proceeding thoughtfully but carefully.